Head of Technology & AI Solutions

I Build AI Systems
That Work While
You Sleep

Multi-agent AI architectures, end-to-end automation, and cloud infrastructure — engineered for US and EU businesses that need real outcomes, not demos.

See What I Build Let's Talk
20+
Clients Served
50+
AI Agents Deployed
10K+
Hours Automated
20%
Lead-to-Close Rate

Stack I build with

What I Build

AI Engineering,
Start to Scale

Multi-Agent AI Systems

Orchestration of Claude, GPT-4, Gemini, and Llama agents running in parallel. Custom LLM pipelines with tool use, memory, and context management — built for production workloads, not demos.

Parallel agent orchestration
LLM pipeline architecture
Tool use, RAG, function calling

Automation & Integrations

n8n, Make.com, and GoHighLevel workflows that eliminate manual ops. CRM automation, webhook pipelines, and API integrations that run 24/7 without you touching them.

n8n & Make.com workflows
GoHighLevel & CRM automation
Webhook & API integrations

Web & Cloud Infrastructure

WordPress, AWS, DNS, and Next.js / React full-stack apps. From technical SEO and landing pages to managed multi-site deployments that stay fast, secure, and online.

AWS, DNS, WordPress management
Next.js & React apps
Technical SEO & performance
Process

From Idea to
Production in Weeks

01

Discovery

We map your operations and identify the highest-leverage automation points — no generic playbook, just what actually moves your business forward.

02

Architecture

System design, model selection, tool stack, and integration plan. Every decision is explained before a single line of code is written.

03

Build & Deploy

Agents coded, tested, and deployed into your stack. Handover includes documentation and a walkthrough so your team can run it independently.

04

Ongoing Support

Monitoring, iteration, and expansion as your needs grow. Systems should improve over time — I stay involved to make sure they do.

Featured Project

AI Portfolio
Screening Platform

The Problem

Recruiters spend hours manually reviewing portfolios. Candidates get no feedback. The process is slow, inconsistent, and impossible to scale.

The System

A multi-agent pipeline where UI, UX, and PM specialist agents evaluate portfolios in parallel, each applying role-specific scoring criteria. Results in seconds, not hours. Two-sided product — recruiters post roles, candidates get scored feedback.

Architecture

01 Portfolio ingestion via Playwright scraper
02 3 specialist agents run in parallel (Claude + OpenAI)
03 Orchestrator aggregates scores into structured JSON
04 Two-sided dashboard for recruiters & candidates

Stack

Claude OpenAI Node.js Playwright TypeScript
About

Ops Instinct Meets
AI Engineering

I'm Yakov (Kobi) Sapir — 16 years in operational leadership, building and managing teams across technology and operations. Three years ago I went deep into AI engineering: production multi-agent systems, cloud infrastructure, full-stack development, and every AI tool worth knowing.

That combination matters. I don't just understand what the tech can do — I understand what a business actually needs. That means fewer wasted sprints, smarter scope decisions, and systems that hold up in production. Past work includes managing 10+ WordPress sites and a full AWS stack for US clients, leading a 5-person technical team, and selling premium AI programs at 15–25% close rates.

Israel — EU timezone
US East Coast overlap
Remote-first
Contact

Got a project in mind?
Let's build it.

Drop me a message or reach out directly — I typically respond within a few hours.

Email

kobisapir0312@gmail.com

WhatsApp / Phone

+972-50-607-5597

LinkedIn

linkedin.com/in/yakov-sapir